The piece emphasizes the importance of distinguishing between different types of R&D investment in software companies, particularly the balance between immediate revenue-generating features and ongoing maintenance and improvements, referred to as "care & feeding." The author argues that while go-to-market executives often prioritize short-term revenue opportunities, this approach can undermine the long-term health of products and technical infrastructure. Care & feeding includes essential tasks like bug fixes, scalability, and UX improvements that keep existing products reliable and competitive. The author suggests approaching R&D budgeting as a portfolio management exercise, similar to financial investments. He proposes allocating resources into four categories: 50% for big revenue drivers (features that are customer-demanded), 35-40% for care & feeding, 5-10% for validation and discovery efforts, and 5% for unplanned or deal-driven work. By framing R&D spending in this way, companies can make better decisions about where to invest and what trade-offs to accept. He warns that if unplanned work exceeds 10%, the organization risks shifting towards a professional services model, which conflicts with the goals of scalable product development. The article also highlights the common pitfalls organizations face. Those that allocate more than 65% of their effort to new features often neglect existing products, leading to a decline in quality and increased technical debt. Conversely, teams focusing excessively on technical maintenance can become stuck in a cycle of constant replatforming without delivering new value to customers. The author uses these observations to encourage leaders to critically assess their R&D spending and ensure a healthier balance that supports both immediate and long-term growth.